forecast_feature_results_multi: Forecasting with FEBAMA

In lily940703/febama: Feature-based Bayesian Forecasting Model Averaging

Forecasting with FEBAMA

Description

Forecast based on FEBAMA framework and obtain error measures.

Usage

forecast_feature_results_multi(
  ts,
  model_conf = model_conf_curr,
  intercept = T,
  data,
  beta_out
)

Arguments

ts	A list with related information of a time series: x Historical data. xx Real data in the forecasting horizon.
model_conf	Parameter settings of FEBAMA framework. Defualt model_conf_default().
data	A list with lpd Log probability densities from historical data. feat Features from historical data. feat_mean The mean of features. feat_sd The sd of features.
beta_out	The output of function febama_mcmc.

Value

forecast_feature_results_multi returns a list with the entries:

err_feature

Log score, mase, smape measures of forecasts.

ff_feature

Forecast values in the forecasting horizon.

w_time_varying

Time-varying weights in the forecasting horizon.

lily940703/febama documentation built on March 20, 2022, 1:57 a.m.